Facial Affect Sensitivity Training for Young Children With Callous-unemotional Traits

The goal of this study is to test a novel intervention for children ages 6-11 with elevated callous-unemotional (CU) traits. Conduct problems are among the most prevalent and costly mental health conditions of childhood, and a common antecedent to adult psychiatric disorders. An established risk factor for early, persistent, and severe youth misconduct is the presence of CU traits. CU traits (e.g., lack of empathy or guilt, shallow affect) are analogous to the core affective features of adult psychopathy, interfere with child socialization, and predict poorer outcomes, even with well-established treatments for disruptive behavior disorders. Thus, novel intervention approaches are needed to target CU traits. Youth with elevated CU traits show deficits in facial emotion recognition (FER) for distress-related expressions, particularly fear or sadness. The central hypothesis is that impaired sensitivity for emotional distress cues (fear and/or sadness) is mechanistically linked to CU traits in children, and that, by targeting affect sensitivity directly, intervention can exert downstream effects on CU traits. A gap in the field regards how to remediate these neurocognitive deficits. This project will directly target affect sensitivity in high-CU youth. The investigators propose an experimental therapeutics approach to develop a novel neurocognitive intervention for CU traits, in which a clearly identified target, facial affect sensitivity (FAS), will be engaged and assessed via primary (distress FER accuracy and/or heightened eye gaze) and secondary (electroencephalograph event-related potential) neurocognitive and behavioral processes. If investigators can demonstrate engagement of the target (FAS) in the initial R61 phase, then in the R33 phase, this finding will be replicated with a new, larger sample, and feasibility and preliminary efficacy of FAST on CU traits will be examined. The long-term goal is to examine FAST impact on behavioral outcomes and to potentially apply this targeted intervention to the wider range of problems associated with CU traits.

This project will directly target affect sensitivity in high-CU youth. Per the National Institute of Mental Health (NIMH) Strategic Plan (Objective 3.1), the investigators propose an experimental therapeutics approach to develop a novel neurocognitive intervention for CU traits, in which a clearly identified target, facial affect sensitivity (FAS), will be engaged and assessed via primary [distress facial emotion recognition (FER) accuracy, heightened eye gaze] and secondary (EEG event-related potential) neurocognitive and behavioral processes. The long-term goal is to apply this targeted intervention to the wider range of problems associated with CU traits. The R61 phase Specific Aims are as follows: Investigators will first demonstrate, in a preliminary randomized controlled trial (RCT; N=84 children), that a new neurocognitive intervention (Facial Affect Sensitivity Training: FAST) can improve FAS [target engagement] in children with elevated CU traits. FAS will be measured primarily by FER accuracy for distress expressions and/or heightened attention to the eye region (eye gaze), and secondarily by neural activity [specifically, N170 and P200 event-related potential (ERP) components]. Objective 1: Establish that distress FER accuracy and/or eye gaze can be altered in a reliable manner among young children with elevated CU traits. Objective 2: Determine whether FAST improves secondary neural indices of FAS (brain activity during processing of emotional faces). Objective 3: Refine FAST for subsequent evaluation by determining optimal dose parameters with regard to number of sessions for FER and/or eye gaze improvement via a nonlinear mixed model for small samples (e.g., timing of local bump or decay, amount of change, when maximal change occurs), and participant satisfaction with session frequency, length, and number. Objective 4: Deliver a computerized training program (FAST) capable of providing real-time automated feedback and reinforcement of accurate FER performance. Milestones (Go/No-Go Criteria): (1) FAST will engage the target (FAS), indexed by enhancing distress FER accuracy and/or eye gaze in high-CU youth. The investigators will examine individual growth rates and test slope differences between conditions (FAST v control). Target engagement will be defined as medium effect size (defined as Cohen's d value = .50) in the comparison of FAST vs. no-treatment control on the primary target (distress FER and/or eye gaze). The R33 phase Specific Aims are as follows: The investigators aim to replicate target engagement with a new, large high-CU sample and evaluate feasibility and preliminary efficacy of FAST, in the context of an RCT (N = 84) in which FAST is compared to an active control condition (ACC; implicit eye gaze training). In addition, this phase will validate the functional role of FAS by examining downstream change in CU as a result of FAST. FAST will produce reliable increases in FER accuracy for distress cues in others. Furthermore, FAST completers will show greater improvement in CU/empathic behaviors than ACC completers. Objective 5: Replicate target engagement of FAS. Objective 6: Determine if improved FAS leads to reduction in CU traits. The investigators will also consider in a preliminary fashion whether CU trait reductions are clinically significant (more than .5 pre-test standard deviation on 2 target CU indices or more than 1.0 standard deviation on one CU measure). If the FAST intervention improves FER and reduces CU traits, such training in early childhood could help interrupt the developmental cascade toward antisocial outcomes.

Attention [F02.830.104.214], Temperament [F01.752.898], Problem Behavior [F01.145.179.750], Social Behavior [F01.145.813], Electroencephalography [E01.370.405.245], Neuropsychological Tests [F04.711.513]

This study utilizes a randomized controlled trial design to test whether a new neurocognitive intervention (Facial Affect Sensitivity Training: FAST) can improve facial affect sensitivity (FAS) in children with elevated CU traits. In FAST, which uses exogenous incentives to train FAS, a series of emotional facial expressions are presented on a computer monitor. Child participants are instructed to look at the eyes of the emotional expressions they view, and must select from a list of emotions the one that matches the facial expression they saw on each trial, which results in reinforcement of correct responses via auditory reward tone. Correct responses also earn points toward incrementally "priced" small toys given at the end of each session. The objective is to make appropriate eye gaze more automatic during FER (thus improving facial affect sensitivity) by repeatedly pairing correct responses with a reward.

An assessment-naïve independent evaluator not involved in intervention implementation will function will provide Clinical Global Impression (CGI) evaluations.

Individuals in this Arm will receive the FAST intervention, as described in the Intervention section of the Clinical Trials form below, with a focus on demonstrating target (facial affect sensitivity) engagement.

Individuals randomized this Arm of the R33 phase will receive the FAST intervention, with the aim of replicating FAST target engagement (as demonstrated in the R61 phase) with a new high-CU sample, and to evaluate the FAST intervention in comparison to an active control condition (Arm 4, implicit eye gaze training).

Individuals in this Arm will receive the active control component, which is an implicit gaze training intervention.

The FAST intervention program represents a novel computerized intervention for high-risk youth that strategically targets implicated facial affect sensitivity deficits directly via a computerized real-time automated feedback and incentive system to remediate callous-unemotional tendencies associated with behavioral dysfunction.

This computerized task was developed to target implicit training of eye gaze but not facial emotion recognition per se via real-time feedback and incentives. On each trial, a fixation cross is followed by an emotional face with eyes directed either left, straight ahead, or right (balanced across expressions), followed by a response key. The child's task is to say which direction the eyes are looking (e.g., "1" or "left"). Stimuli are black and white standardized photographs of men and women models from the Ekman Pictures of Facial Affect each displaying the 3 gaze directions for 6 emotion expressions.

Recognition of facial emotional expressions will be indexed based on accuracy of matching emotion expressions on the FACES and Dynamic FER tasks.

FER will be assessed in each phase (R61 & R33) of the study at pre-treatment (session 1), post-treatment (end of 5 weeks), and 3-month follow-up as well as approximately every other week during the 5-week intervention.

Heightened attention to the eye region (eye gaze) will be indexed by primacy, dwell time, and frequency of fixation on the eye region of emotional faces, measured via eye tracking while participants complete the FER task.

Eye gaze will be assessed in each phase (R61 & R33) of the study at pre-treatment (session 1), post-treatment (end of 5 weeks), and 3-month follow-up as well as approximately every other week during the 5-week intervention.

CU traits will be assessed using the Inventory of Callous-Unemotional Traits. There are subscales for callousness, uncaring, and unemotional tendencies. Higher scores reflect higher reported CU traits.

CU Traits will be assessed during the the R33 phase at pre-treatment, post-treatment (end of 5 weeks), and 3-month follow-up as well as approximately every other week during the 5-week intervention.

Griffith Empathy Measure (GEM ), is a 23-item parent-report measure of affective (shared emotional state) and cognitive (perspective taking) empathy. The maximum score is 92. Higher scores reflect higher reported empathy.

GEM will be assessed during the R33 phase at pre-treatment, post-treatment (end of 5 weeks), and 3-month follow up, and approximately every other week during the 5-week intervention.

The I Love You Task is a dyadic interactional task for young children utilizing a brief emotionally intense parent-child encounter for which reciprocated eye gaze and affection is fundamental.

The "I Love You" task will be administered during the R33 phase at pre-treatment, post-treatment (end of 5 weeks), and 3-month follow up

The Clinical Global Impression (CGI) score will be determined as a common clinical trial index, with CGI-I improvement scores of 1 or 2 (very much/much improved, based on independent rater score) as being responsive to the intervention.

CGI will be assessed at eligibility, post-treatment (end of 5 weeks), and 3-month follow-up in the R33 phase.

Event related potentials that are previously identified neural correlates of early perceptual and emotional processing of facial expressions, specifically, N170 and P200 ERP components, assessed via electroencephalography.

ERPs will be assessed at eligibility, post-treatment (end of 5 weeks), and 3-month follow-up in both R61 and R33 phases.

Inclusion Criteria: A standard score less than or equal to 8 on the NEPSY (A Developmental NEuroPSYchological Assessment) Affect Recognition (AR) test, or less than or equal to 70% accuracy for distress-related emotions on a Dynamic FER measure. Composite intelligence quotient (IQ) score of at least 80 on the Wechsler Abbreviated Scale of Intelligence, Second Edition. Any psychotropic medications must be on stable dosing schedule for 2 weeks prior to entry. Presence of elevated CU traits (defined as in prior studies as score of "2" on at least 2 of the 4 CU items on the Antisocial Process Screening Device (APSD). Exclusion Criteria: Bipolar disorder. Current risk for suicide or harm to others. Autism spectrum disorder (ASD). Currently participating in therapy for CU traits or facial emotion recognition deficits.

Per the Dissemination Plan, the investigators will follow the NIH Policy on Dissemination of NIH-Funded Clinical Trial Information and University of Alabama's Office for Sponsored Research policy of compliance monitoring (Sponsored Projects Agreements for Human Subjects Research and Human Research Protections).

The investigators will make de-identified data generated through this project available for research purposes to qualified individuals within the scientific community no later than the time of publication of the main findings from the final data set. Additionally, data will be shared with the NIMH Data Archive (NDA) that includes the National Database for Clinical Trials related to Mental Illness (NDCT). The investigators will abide by the data submission schedule given us in the NDA. Data will be provided within the recommended data structures of the NDA. In addition, the study will be registered with ClinicalTrials.gov and will share results with ClinicalTrials.gov within one year of the completion of data collection of the primary outcome measures. Sharing of the data generated by this project will be in accordance with all applicable regulations and guidelines of NIH.

Requests for individual participant data (IPD) or other study information will be reviewed by the PI (B. White). He will make determination as to legitimacy of the request and qualifications of the requestor.